CN105137969A - Quadruped robot trot gait and body gesture control method based on support line motion decomposition - Google Patents

Abstract

The invention discloses a quadruped robot trot gait and body gesture control method based on support line motion decomposition, and the method comprises the steps: S1, enabling a robot to be respectively projected to a radial plane and a normal plane, wherein the radial plane passes through a supporting leg foot end point line and is perpendicular to a horizontal plane, and the normal plane passes through a mass center of the body and is perpendicular to the radial plane; S2, enabling an expected speed of the robot body to be projected to the radial plane and the normal plane through employing the same projection method as the step S1, and obtaining the expected speeds of the body on the two projection planes; S3, respectively building a kinetic and dynamic equation of a simplified model, obtaining the body speeds on the two projection planes to follow the respective expected track through employing a controller, carrying out speed synthesis, and completing the control of the speed and gesture of the robot body. The method is high in universality, is good in control effect, and is high in control precision.

Description

Based on the quadruped robot trot gait body posture control method of Support Level Kinematic Decomposition

Technical field

The present invention is mainly concerned with motion planning and robot control technical field, refers in particular to a kind of quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition.

Background technology

Quadruped robot is the skeletal structure of quadruped mammal and the robot of walking step state in a kind of natural imitation circle, fully combine the advanced technology of bionics and robotics, its maximum advantage has extremely strong adaptive faculty to non-structured terrain environments such as mountain region, abrupt slope, deserts, may be used for performing patrol under hazardous environment, some complex tasks such as Material Transportation, gather around and have broad application prospects.

Quadruped robot is generally made up of a body and four bionic legs, and every bar bionic leg is made up of 2 ~ 3 forward direction joints and a side direction joint, to ensure that the motion of sufficient end points has three degree of freedom.The joint driver of the quadruped robot of current comparative maturity is divided into two kinds: hydraulic unit driver and motor driver.By the study of the mode of motion to quadruped mammal, analysis, imitation, the movement of quadruped robot mainly adopts following three kinds of typical gait: WALK gait (Crawl gait), TROT gait (trot gait) and BOUNDING gait (gait of running).

At quadruped robot with in TROT gait traveling process, the motion of diagonal angle two legs is identical, and robot relies on two groups of diagonal angle legs to switch in supporting leg pattern and periodicity between pattern of leading leg according to certain rule, completes the control to body speed and attitude.At present, quadruped robot motion control generally adopts following methods:

1, CPG control method; I.e. central pattern generator (cpg), produces cyclical signal by neural oscillator model, controls leg exercise, and the advantage of this method realizes simply, and shortcoming is that parameter is many, and lacks clear and definite physical meaning, is difficult to accurate adjustment.

2, machine learning; By the Theories and methods of machine learning, enable robot learn the stable walking step state of quadruped, but need a large amount of training, be difficult to adapt to new unstructured moving grids in the short time.

3, plan online; The sensor information current according to robot, calculates next or several walking period foot end points desired motion track, makes sufficient end points follow the tracks of desired motion track by adjustment joint driven torque.But due to the non-intellectual of unstructured moving grids, situation about contacting in advance with environment of leading leg may be there is, produce unexpected impulsive force, affect speed and the attitude of body.

4, sufficient termination touch controls; By force snesor feedback foot end and the contact force information of environment, design sufficient termination touch control algolithm, and then the speed of control machine human body and attitude.The shortcoming of this control method is may produce internal force between two supporting legs, when supporting leg be switched to lead leg pattern when, internal force abrupt release, to lead leg and the attitude of body produces unexpected impact.Therefore, need further design con-trol strategy, can either the speed of control machine human body and attitude, effectively can eliminate again the internal force between supporting leg.

Summary of the invention

The technical problem to be solved in the present invention is just: the technical matters existed for prior art, the invention provides a kind of highly versatile, control effects is good, control accuracy the is high quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition.

For solving the problems of the technologies described above, the present invention by the following technical solutions:

Based on a quadruped robot trot gait body posture control method for Support Level Kinematic Decomposition, the steps include:

S1: robot is projected to respectively on sagittal plane and Normal plane; Described sagittal plane be by supporting leg foot end points line and with the plane of horizontal plane; Described Normal plane is by body barycenter and perpendicular to the plane of sagittal plane;

S2: use the method identical with step S1 to project on sagittal plane and Normal plane the desired speed of robot body, obtain the body desired speed on two projection planes;

S3: kinematics and the kinetics equation of setting up simplified model respectively, by the body speed tracing desired trajectory separately adopting controller to make on two projection planes, then carries out velocity composite, completes the control of speed to robot body and attitude.

As a further improvement on the present invention: in described step S1 and S2, the projection on described sagittal plane is reduced to plane seven-link assembly model; Centroid position, the Attitude Tracking desired trajectory of plane seven-link assembly is made by adjustment joint driven torque.

As a further improvement on the present invention: in described step S1 and S2, the projection on described Normal plane is reduced to the linear inverted pendulum model of one dimension.

As a further improvement on the present invention: in described step S3, by the position of foot point of leading leg in calculating, selection Normal plane, adjust speed, the attitude of each start time support phase robot body, control the speed of body in support finish time phase Normal plane, synthesize with the speed of sagittal plane inner machine human body, complete the control to robot speed.

As a further improvement on the present invention: adopt the dicyclo PID controller controlled based on sufficient termination touch, outer shroud utilizes the sufficient termination touch of body barycenter and attitude error calculation expectation, and add supporting leg foot end between internal force be the constraint condition of zero, calculate unique expectation contact force; Inner ring is by measuring actual sufficient termination touch and expecting the error calculation joint driven torque between contact force.

Compared with prior art, the invention has the advantages that:

1, the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition of the present invention, at quadruped robot with in TROT gait traveling process, the velocity gesture adopting the control strategy based on resolution of velocity and synthesis can realize body accurately follows the tracks of desired trajectory.

2, the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition of the present invention, the control strategy proposed can effectively reduce the internal force of quadruped robot in the process of walking between supporting leg foot end.

3, the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition of the present invention, can improve the adaptive faculty of quadruped robot to unstructured moving grids, realize robot in the ground walking of out-of-flatness.

4, the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition of the present invention, has stronger versatility.Control strategy does not rely on concrete quadruped robot system, as long as build suitable sensing system, in conjunction with robot kinematics, kinetic model correlation parameter, suitably adjusts controller parameter, can realize method proposed by the invention.

5, the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition of the present invention, clear in structure, well arranged, there is good Theory and applications and be worth.

Accompanying drawing explanation

Fig. 1 is the system architecture schematic diagram of quadruped robot.

Fig. 2 is the sagittal plane perspective view of the present invention's quadruped robot when embody rule.

Fig. 3 is the Normal plane perspective view of the present invention's quadruped robot when embody rule.

Fig. 4 is the control flow schematic diagram of the present invention in embody rule example.

Fig. 5 is the concrete computation process schematic diagram of the present invention in embody rule example.

Fig. 6 is the schematic flow sheet of the inventive method.

Embodiment

Below with reference to Figure of description and specific embodiment, the present invention is described in further details.

Method of the present invention is mainly applicable to quadruped robot, and as shown in Figure 1, it comprises a body and four bionic legs to the system architecture of quadruped robot, namely comprises left front leg, RAT, left back leg, right rear leg.Article four, the physical construction, measure-alike of bionic leg, left front leg and left back leg, the mounting means of RAT and right rear leg is relative to body mirror image each other.Every bar bionic leg comprises in four initiatively joint (ankle-joint, knee joint, hip joint, hip side direction joints), and each initiatively joint all adopts hydraulic actuator to drive, and foot end is point cantact with ground.Displacement transducer and force snesor are equipped with in each active joint, for detecting actuator length information and joint force information, at sufficient end points by the contact force information of three-dimensional force sensor feedback foot end with ground, the position of robot body in inertial space, attitude are detected in real time by the IMU attitude sensor being arranged on upper body.

In the present invention, walking step state is defaulted as TROT gait.As shown in Figure 6, the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition of the present invention, the steps include:

S1: robot is projected to respectively on sagittal plane and Normal plane;

Described sagittal plane be by supporting leg foot end points line and with the plane of horizontal plane; Described Normal plane is by body barycenter and perpendicular to the plane of sagittal plane;

Projection on described sagittal plane is reduced to plane seven-link assembly model (as shown in Figure 2), and the projection on described Normal plane is reduced to the linear inverted pendulum model of one dimension (as shown in Figure 3);

S2: use the method identical with step S1 to project on sagittal plane and Normal plane the desired speed of robot body, obtain the body desired speed on two projection planes;

Based on the expectation forward speed V of body _d(the side direction desired speed of body is 0), calculates the desired speed of body in two projection planes.

V _sd＝V _dcosα

V _fd＝V _dsinα

Wherein V _dfor the forward direction desired speed of body, V _sdfor the desired speed of body in sagittal plane, V _fdfor the desired speed of body in Normal plane, α is the angle between robot forward speed and sagittal plane.

S3: kinematics and the kinetics equation of setting up simplified model respectively, by the body speed tracing desired trajectory separately adopting controller to make on two projection planes, then carries out velocity composite, to complete the control of speed to robot body and attitude.

Be in the support phase with left front leg and right rear leg, it is example that RAT and left back leg are in shaking peroid:

The kinetic model of TROT gait quadruped robot is:

Sagittal plane:

${\tilde{M}}_1{\stackrel{\·\·}{q}}_1+{\tilde{C}}_1{\stackrel{\·}{q}}_1+{\tilde{N}}_1={\tilde{J}}_1{F}_e$

$M_2{\stackrel{\·\·}{q}}_2+C_2{\stackrel{\·}{q}}_2+N_2=\mathrm{\τ}+J_2{F}_e$

Wherein, q ₁=[X, Y, β] ^trepresent the horizontal level of body in sagittal plane, height and the angle of pitch,

Q ₂=[X _c, Y _c, θ _d1, X _d, Y _d, θ _d2] ^t, [X _c, Y _c] ^t([X _d, Y _d] ^t) for right rear leg (left front leg) foot end is relative to the position of hip joint mounting points, θ _d1(θ _d2) be the linear restriction of right rear leg (left front leg) forward direction joint angle, F _e=[F _xe1, F _ye1, F _xe2, F _ye2] ^tfor supporting leg foot end in contact force vector, τ is supporting leg joint driven torque, for inertial matrix, for coriolis force matrix, for gravity item, for Jacobian matrix.

Normal plane:

$l\left(t\right)=l\left(0\right)\cosh (t/T_c)+T_c\stackrel{\·}{l}\left(0\right)\sinh (t/T_c)$

$\stackrel{\·}{l}\left(t\right)=l\left(0\right)\sinh (t/T_c)/T_c+\stackrel{\·}{l}\left(0\right)\cosh (t/T_c)$

$T_c=\sqrt{Y/g}$

Wherein, l (t) is the horizontal level of body in Normal plane, and l (0) is the horizontal level of support phase initial time body in Normal plane, for supporting the horizontal velocity of phase initial time body in Normal plane.G is acceleration of gravity.

Controller is divided into inner ring and outer shroud two parts:

Outer ring controller:

$F_{ed}={\left({\tilde{J}}_1\right)}^{-1}\left({\tilde{M}}_1\left({\stackrel{\·\·}{q}}_{1d}+{\mathrm{Kp}}_1\left(q_{1d}-q_1\right)+{\mathrm{Kd}}_1\left({\stackrel{\·}{q}}_{1d}-{\stackrel{\·}{q}}_1\right)+{\mathrm{Ki}}_1{\∫}_0^T\left(q_{1d}-q_1\right)dt\right)+{\tilde{C}}_1{\stackrel{\·}{q}}_1+{\tilde{N}}_1\right)$

Wherein q _1dfor q ₁desired trajectory, F _edfor supporting leg foot end expects contact force, Kp ₁, Kd ₁, Ki ₁for outer shroud controling parameters.

Interior ring controller:

$\mathrm{\τ}=-(M_2{\stackrel{\·\·}{q}}_2+C_2{\stackrel{\·}{q}}_2+N_2)+J_2\left({\mathrm{Kp}}_2\right(F_{ed}-F_e\left)\right)$

Wherein, Kp ₂for inner ring controling parameters.

T ₁the moment computing formula of linear inverted pendulum initial position is:

$l\left(t_1^{+}\right)=\left(V_{fd}^{-}\left(t_1+T\right)-V\left(t_1\right)\sin \mathrm{\α}\cosh \left(T/T_c\right)\right)T_c/\sinh (T/T_c)$

Wherein, T is single time support phase.

In above process, by robot projection on sagittal plane, after being reduced to a plane seven-link assembly, made centroid position (2 dimension), attitude (1 dimension) the tracking desired trajectory of plane seven-link assembly by adjustment joint driven torque.

In above-mentioned steps S3, the present invention adopts the dicyclo PID controller controlled based on sufficient termination touch, outer shroud utilizes the sufficient termination touch of body barycenter and attitude error calculation expectation, because controlled quentity controlled variable is 4 dimensions, control objectives is 3 dimensions, so adding internal force between supporting leg foot end is the constraint condition of zero, calculate unique expectation contact force.Inner ring is by measuring actual sufficient termination touch and expecting the error calculation joint driven torque between contact force.Further, if forward direction joint number is greater than 2, also add suitable constraint condition, to ensure the uniqueness of joint driven torque.

Because the height of body is controlled in sagittal plane, so the projection of robot in Normal plane can be reduced to an one-dimensional linear inverted pendulum (see Fig. 3).When gait cycle is fixed, when robot body side direction desired speed is 0, according to the kinematics and dynamics modeling of one dimension inverted pendulum, each finish time support phase the forward speed of robot only support the initial velocity of start time phase robot body therewith, attitude is correlated with.

For this reason, the present invention is by the position of foot point of leading leg in calculating, selection Normal plane, ensureing under robot stabilized prerequisite, adjust speed, the attitude of each start time support phase robot body, and then control the speed of body in support finish time phase Normal plane, synthesize with the speed of sagittal plane inner machine human body, complete the control to robot speed.

In above-mentioned steps S1 and step S2, when carrying out modeling and control in sagittal plane, comprise following process:

The projection of quadruped robot in sagittal plane can be reduced to a seven-link assembly structure, (supposes that left front leg and right rear leg are as supporting leg) as shown in Figure 2.(X, Y) is the position of body barycenter under inertial coordinates system, L ₀for robot body length, d is the width of body, L ₁, L ₂, L ₃be respectively the length of leg ankle-joint, knee joint, hip joint, θ ₁, θ ₂, θ ₃for left front leg ankle-joint angle, knee angle, hip joint angle, θ ₄, θ ₅, θ ₆for right rear leg ankle-joint angle, knee angle, hip joint angle, β is the body angle of pitch, O, A, B, C, D are respectively body coordinate system initial point (i.e. body barycenter), right rear leg hip and body tie point, left front leg hip and body tie point, right rear leg foot end points, left front leg foot end points, F _xe1, F _ye1, F _xe2, F _ye2be respectively the contact force that C point and D point are subject to.The position coordinates of C point and D point is:

X _C＝(X-(L ₀/2)cosβ)-L ₃cos(θ ₆-β)+L ₂cos(θ ₅-θ ₆+β)-L ₁cos(θ ₄-θ ₅+θ ₆-β)

(1)

Y _C＝(Y-(L ₀/2)sinβ)-L ₃sin(θ ₆-β)-L ₂sin(θ ₅-θ ₆+β)-L ₁sin(θ ₄-θ ₅+θ ₆-β)

X _D＝(X+(L ₀/2)cosβ)+L ₃cos(θ ₃-β)-L ₂cos(θ ₂-θ ₃+β)+L ₁cos(θ ₁-θ ₂+θ ₃-β)

(2)

Y _D＝(Y+(L ₀/2)sinβ)-L ₃sin(θ ₃-β)-L ₂sin(θ ₂-θ ₃+β)-L ₁sin(θ ₁-θ ₂+θ ₃-β)

Based on Lagrange's equation, the kinetic model of seven-link assembly structure is:

$M_1{\stackrel{\·\·}{q}}_1+C_1{\stackrel{\·}{q}}_1+N_1=J_1{F}_e$

(3)

$M_2{\stackrel{\·\·}{q}}_2+C_2{\stackrel{\·}{q}}_2+N_2=\mathrm{\τ}+J_2{F}_e$

Wherein, q ₁=[X, Y, β] ^t, q ₂=[X _c, Y _c, θ _d1, X _d, Y _d, θ _d2] ^t, F _e=[F _xe1, F _ye1, F _xe2, F _ye2] ^t, τ is for corresponding to θ ₁... θ ₆joint driven torque, M ₁, M ₂for inertial matrix, C ₁, C ₂for coriolis force matrix, N ₁, N ₂for gravity item, J ₁, J ₂for Jacobian matrix.

J in formula (3) ₁for 3*4 Jacobian matrix, directly can not ask for inverse matrix, the present invention is for the consideration of compliance, and adding internal force between supporting leg foot end points is the constraint condition of 0, that is:

$\left[\begin{array}{c}{F}_{xe1}-F_{xe2}\\ F_{ye1}-F_{ye2}\end{array}\right]\·\left[\begin{array}{c}{X}_D-X_C\\ Y_D-Y_{\∂C}\end{array}\right]=0---\left(4\right)$

According to formula (3) and (4), the kinetic model of seven-link assembly system becomes:

${\tilde{M}}_1{\stackrel{\·\·}{q}}_1+{\tilde{C}}_1{\stackrel{\·}{q}}_1+{\tilde{N}}_1={\tilde{J}}_1{F}_e$

(5)

$M_2{\stackrel{\·\·}{q}}_2+C_2{\stackrel{\·}{q}}_2+N_2=\mathrm{\τ}+J_2{F}_e$

Wherein, for 4*4 Jacobian matrix.

According to the kinetic model of seven-link assembly system, the dicyclo PID controller based on power inner ring of the present invention:

Outer ring controller:

$F_{ed}={\left({\tilde{J}}_1\right)}^{-1}\left({\tilde{M}}_1\right({\stackrel{\·\·}{q}}_{1d}+{\mathrm{Kp}}_1(q_{1d}-q_1)+{\mathrm{Kd}}_1({\stackrel{\·}{q}}_{1d}-{\stackrel{\·}{q}}_1)+{\mathrm{Ki}}_1{\∫}_0^T(q_{1d}-q_1)dt)+{\tilde{C}}_1{\stackrel{\·}{q}}_1+{\tilde{N}}_1)---\left(6\right)$

Wherein q _1dfor q ₁desired trajectory, F _edfor supporting leg foot end expects contact force.

Interior ring controller:

$\mathrm{\τ}=-(M_2{\stackrel{\·\·}{q}}_2+C_2{\stackrel{\·}{q}}_2+N_2)+J_2\left({\mathrm{Kp}}_2\right(F_{ed}-F_e\left)\right)---\left(7\right)$

In above-mentioned steps S1 and step S2, in Normal plane during modeling and control, comprise following process:

Because the height of robot body is controlled in sagittal plane, so the projection of quadruped robot in Normal plane can be reduced to a linear inverted pendulum (as shown in Figure 3).The motion model of linear inverted pendulum is as follows:

$l\left(t\right)=l\left(0\right)\cosh (t/T_c)+T_c\stackrel{\·}{l}\left(0\right)\sinh (t/T_c)$

$\stackrel{\·}{l}\left(t\right)=l\left(0\right)\sinh (t/T_c)/T_c+\stackrel{\·}{l}\left(0\right)\cosh (t/T_c)---\left(8\right)$

$T_c=\sqrt{Y/g}$

As can be seen from formula (8), the projection of speed on Normal plane supporting finish time phase robot body just and the position of this support start time phase body and velocity correlation.In the present invention, the control strategy in Normal plane to be led leg the position of foot point by adjustment, thus change position and the speed of next start time support phase body, and then control the speed of finish time support phase body.Concrete computation process in a gait cycle (supposes that current time is t as follows ₀, next start time support phase is t ₁, supporting time phase is T, and the angle of supporting leg foot end points line and robot working direction is α, and current supporting leg is left front leg and right rear leg, t ₁-t ₁+ T time inner support leg is RAT and left back leg, t ₁+ T-t ₁+ 2T time inner support leg is left front leg and right rear leg, and as shown in Figure 4, concrete solution process is as shown in Figure 5 for process flow diagram):

The first step: calculate t ₁the desired speed of+T moment body in sagittal plane and Normal plane

$V_{sd}^{-}(t_1+T)=V_d(t_1+T)cos\mathrm{\α}$

(9)

$V_{fd}^{-}(t_1+T)=V_d(t_1+T)sin\mathrm{\α}$

V _dt (), V (t) are t body forward direction desired speed, actual speed, subscript ^-( ⁺) represent according to (afterwards) supporting leg foot end line angle decomposition sagittal plane and Normal plane before the foot that falls.

Second step: based on formula (8), according to calculate moment linear inverted pendulum initial position:

$l\left(t_1^{+}\right)=\left(V_{fd}^{-}\right(t_1+T)-V_f^{+}(t_1\left)\cosh \right(T/T_c\left)\right)T_c/\sin h(T/T_c)---\left(10\right)$

Wherein, $V_f^{+}\left(t_1\right)=V\left(t_1\right)sin\mathrm{\α}.$

3rd step: speed trajectory in speed trajectory, Normal plane in calculating body x direction speed trajectory, sagittal plane:

$V_f\left(t\right)=l\left(t_1^{+}\right)\sinh \left(\right(t-t_1)/T_c)/T_c+V_f^{+}\left(t_1\right)\cosh \left(\right(t-t_1)/T_c),t\∈\[t_1^{+},t_1^{-}+T\]---\left(11\right)$

$V\left(t\right)=V_f\left(t\right)/sin\mathrm{\α},t\∈\[t_1^{+},t_1^{-}+T\]---\left(12\right)$

$V_s\left(t\right)=V\left(t\right)cos\mathrm{\α},t\∈\[t_1^{+},t_1^{-}+T\]---\left(13\right)$

V _st the control method of () is described in modeling and control procedure in above-mentioned sagittal plane.

4th step: calculate moment leads leg Luo Zu position, operating space:

RAT falls position, foot point operating space:

$x_{eRF}\left(t_1^{-}\right)=\frac{{\∫}_{t_1^{+}}^{t_1^{-}+T}V\left(t\right)dt}{2}---\left(14\right)$

$z_{eRF}\left(t_1^{-}\right)=tan\mathrm{\α}(\frac{L_0}{2}+x_{eRF}(t_1^{-})-l\left(t_1^{+}\right)/sin\mathrm{\α})-d---\left(15\right)$

Left back leg falls position, foot point operating space:

$x_{eLR}\left(t_1^{-}\right)=\frac{{\∫}_{t_1^{+}}^{t_1^{-}+T}V\left(t\right)dt}{2}---\left(16\right)$

$z_{eLR}\left(t_1^{-}\right)=d-l\left(t_1^{+}\right)/cos\mathrm{\α}-\left(\right(L_0/2-x_{eLR}\left(t_1^{-}\right)\left)tan\mathrm{\α}\right)---\left(17\right)$

5th step: calculate t ₁the desired speed of+2T moment body in sagittal plane and Normal plane

$V_{sd}^{-}(t_1+2T)=V_d(t_1+2T)cos\mathrm{\α}$

(18)

$V_{fd}^{-}(t_1+2T)=V_d(t_1+2T)sin\mathrm{\α}$

6th step: based on formula (8), according to calculate moment linear inverted pendulum initial position:

$l\left(t_1^{+}+T\right)=\left(V_{fd}^{-}\right(t_1+2T)-V_f^{+}(t_1+T\left)\cosh \right(T/T_c\left)\right)T_c/\sinh (T/T_c)---\left(19\right)$

Wherein, $V_f^{+}(t_1+T)=V(t_1+T)sin\mathrm{\α}.$

7th step: speed trajectory in speed trajectory, Normal plane in calculating body x direction speed trajectory, sagittal plane:

$V_f\left(t\right)=l\left(t_1^{+}\right)\sinh \left(\right(t-t_1-T)/T_c)/T_c+V_f^{+}\left(t_1\right)\cosh \left(\right(t-t_1-T)/T_c),t\∈\[t_1^{+}+T,t_1^{-}+2T\]---\left(20\right)$

$V\left(t\right)=V_f\left(t\right)/sin\mathrm{\α},t\∈\[t_1^{+}+T,t_1^{-}+2T\]---\left(21\right)$

$V_s\left(t\right)=V\left(t\right)cos\mathrm{\α},t\∈\[t_1^{+}+T,t_1^{-}+2T\]---\left(22\right)$

V _st the control method of () is identical with the 3rd step.

8th step: calculate moment leads leg Luo Zu position, operating space:

Left front leg falls position, foot point operating space:

$x_{eLF}(t_1^{-}+T)=\frac{{\∫}_{t_1^{+}+T}^{t_1^{-}+2T}V\left(t\right)dt}{2}---\left(23\right)$

$z_{eLF}(t_1^{-}+T)=d-tan\mathrm{\α}(\frac{L_0}{2}+x_{eLF}(t_1^{-}+T)-l\left(t_1^{+}+T\right)/sin\mathrm{\α})---\left(24\right)$

Right rear leg falls position, foot point operating space:

$x_{eRR}(t_1^{-}+T)=\frac{{\∫}_{t_1^{+}+T}^{t_1^{-}+2T}V\left(t\right)dt}{2}---\left(25\right)$

$z_{eLR}(t_1^{-}+T)=-d+l\left(t_1^{+}\right)/cos\mathrm{\α}+\left(\right(L_0/2-x_{eLR}\left(t_1^{-}+T\right)\left)tan\mathrm{\α}\right)---\left(26\right)$

Below be only the preferred embodiment of the present invention, protection scope of the present invention be not only confined to above-described embodiment, all technical schemes belonged under thinking of the present invention all belong to protection scope of the present invention.It should be pointed out that for those skilled in the art, some improvements and modifications without departing from the principles of the present invention, should be considered as protection scope of the present invention.

Claims (5)

1., based on a quadruped robot trot gait body posture control method for Support Level Kinematic Decomposition, it is characterized in that, step is:

S1: robot is projected to respectively on sagittal plane and Normal plane; Described sagittal plane be by supporting leg foot end points line and with the plane of horizontal plane; Described Normal plane is by body barycenter and perpendicular to the plane of sagittal plane;

S2: use the method identical with step S1 to project on sagittal plane and Normal plane the desired speed of robot body, obtain the body desired speed on two projection planes;

S3: kinematics and the kinetics equation of setting up simplified model respectively, by the body speed tracing desired trajectory separately adopting controller to make on two projection planes, then carries out velocity composite, completes the control of speed to robot body and attitude.

2. the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition according to claim 1, it is characterized in that, in described step S1 and S2, the projection on described sagittal plane is reduced to plane seven-link assembly model; Centroid position, the Attitude Tracking desired trajectory of plane seven-link assembly is made by adjustment joint driven torque.

3. the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition according to claim 2, it is characterized in that, in described step S1 and S2, the projection on described Normal plane is reduced to the linear inverted pendulum model of one dimension.

4. the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition according to claim 1 or 2 or 3, it is characterized in that, in described step S3, by the position of foot point of leading leg in calculating, selection Normal plane, adjust speed, the attitude of each start time support phase robot body, controlling body supporting the speed in finish time phase Normal plane, synthesizing with the speed of sagittal plane inner machine human body, completing the control to robot speed.

5. the quadruped robot trot gait body posture control method based on Support Level Kinematic Decomposition according to claim 1 or 2 or 3, it is characterized in that, adopt the dicyclo PID controller controlled based on sufficient termination touch, outer shroud utilizes the sufficient termination touch of body barycenter and attitude error calculation expectation, and add supporting leg foot end between internal force be the constraint condition of zero, calculate unique expectation contact force; Inner ring is by measuring actual sufficient termination touch and expecting the error calculation joint driven torque between contact force.